- Email: [email protected]
Application of ultrasound in comminution
Ultrasonics 38 (2000) 345–352 www.elsevier.nl/locate/ultras
Application of ultrasound in comminution L.F. Gaete-Garreto´n *, Y.P. Vargas-Herma´ndez, C. Velasquez-Lambert Laboratorio de Ultrasonidos, Depto. de Fı´sica USACH, Casilla 307, Santiago-2, Chile
Abstract This paper deals with a new technology for fine grinding of hard materials, based on a high-compression roller mill with ultrasonic capabilities. The machine was tested by producing fine powders from hard rocks, with and without ultrasonic activation, permitting the beneficial effects of ultrasound to be evaluated. The experimental set-up allows the following operational parameters to be measured: material flow, applied torque, angular velocity of the rollers, stress on the shafts, ultrasonic energy applied and the vibration amplitude and phase behaviour of the transducer roller. It is found that the application of ultrasonic energy diminishes the torque required, the stress over the shafts and the total energy consumed for the same grinding results. In addition, a reduction in the erosion of the grinding surfaces was found. The optimal value of the applied ultrasonic power was determined by measuring the specific rate of breakage, a parameter that refers to the energy consumed for the generation of 1 ton of material, for each size range. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Comminution; Grinding; High-compression ultrasonic mill; Size reduction
1. Introduction Comminution is defined as the mechanical breakdown of solids into smaller particles without changing the state of aggregation [1]. This process may be divided into two steps: crushing, which is the reduction of large pieces of material to a size suitable for grinding, and grinding itself, the reduction of crushed material to a powder. In order to produce a mechanical fracture, the material must be stressed beyond a certain critical limit. At present, three main stress mechanisms for particle size reductions are generally used: $ stress applied between two surfaces, applying normal and tangential forces to a single or many particles in the bulk material as in various crushers ( jaw crushers, overhead eccentric jaw crushers, impact crushers, giratory crushers, disk crushers, roll crushers, hammer crushers roller and ring-roller mills); $ stress applied at a single solid surface (surface– particle or particle–particle) such as in grinding media
* Corresponding author. Tel.: +56-2-7797-067; fax: +56-2-7797-067. E-mail address: [email protected] (L.F. Gaete-Garreto´n)
mills (tumbling mills, vibration mills, planetary mills, autogenous mills); $ stress applied by a carrier medium, such as in wet grinding. In conventional grinding machines, the energy that is actually needed for particle reduction is approximately 1% of the total energy applied to the machine. Despite the economic importance of comminution, its technology and efficiency have basically remained unchanged for many years, leading to a great demand for improvements [2,3]. Recent developments include the high-compression roller mill [4], a more efficient machine that introduces several advantages over the classical roller mill. It provides a better performance and is capable of achieving fine grindings from soft materials but is not useful for grinding hard materials due to excessive mechanical wear. The low efficiency of current grinding technologies is mainly due to the application of the stress to the whole material volume. This produces an excess of elastic energy in the load, an important part of which is finally dissipated as heat. Therefore, to increase the grinding process efficiency, it would be advisable to look for a way to apply the grinding energy only to the specific material to be fractured.
0041-624X/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S0 0 4 1 -6 2 4 X ( 9 9 ) 0 0 17 0 - 5
346
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
Some mechanisms to supply the breaking energy just in the grinding material have been explored previously. An example is the Snyder process, which uses a pressure discharge in a chamber filled with coarse material to produce fluidization and a grinding effect. Thermal stress, electrical discharge and electro-hydraulic effects have also been tried as alternative non-impact grinding methods, but have resulted in energy demands an order of magnitude higher than those needed by traditional technology to obtain the same product [5,6 ]. In the meeting of International Comminution Research Association, in Warsaw in 1993, the feasibility of the application of ultrasound energy to the grinding process was seen a viable avenue of exploration and research. Several different reasons were argued in the meeting to support such a proposition. One of the most convincing originated in the accepted fact that inside any material there are a number of inherent cracks and ultrasonic energy has the capacity to produce crack propagation from within the particle to its outer surface, producing an efficient fracture. In addition, in the case of mineral grinding, this method would be qualitatively better than the application of external stress, because the fractures would occur in the natural boundaries within the material compounds, precluding inter-crystalline breakage, a phenomenon that produces new substances, complicating the extraction process. In synthesis, the objective is the efficient transfer of ultrasonic energy from the source to the material to be ground. The fragmentation effect of ultrasonic fields in suspensions and in solids has been investigated since the fifties. Ga¨rtner [7] was probably the first researcher to have attempted using ultrasonic waves in the fragmentation of particles, obtaining poor results. Fleischhauer [8] used ultrasonic waves to fragment coal in suspension, while Graff [9] arranged many transducers in a multistage stamp mill. Quantitative information on the success of these attempts is very scarce. Leach [10] studied the fragmentation of resonant rocks samples fixed to the tip of an ultrasonic transducer, observing a preferred fracture at the nodes. Tarpley and Moulder [11] studied the effect of ultrasound in the grinding of coal under compression. These researchers designed two different machines. The first consisted of a nip roll mill that squeezes the coal grains between a roll and an ultrasonic active plate. Their reported values of energy consumption were as low as 3 kWh/ton to obtain a product 80% under 125 mm, much lower than 20 kWh/ton usually spent in hammer mills. The second machine was a conical device, activated through two opposing ultrasonic transducers with an Auger spiral to control the flux of material and to squeeze the coal against the ultrasonically active conical wall. According to the authors, this device offered a
production capacity of 136 kg h with an energy consumption of 5–7 kWh/ton. Later, in 1988 [12], using the same device, other researchers were unable to reproduce the previously reported results. In 1992, Lo and Kientzler [13] evaluated the Tarpley nip roller machine, using the device to grind mineral specimens. Their results showed energy consumptions similar to that of a ball mill, but not better. Menacho et al. [14] carried out grinding tests comparing standard mineral samples with ultrasonic pre-treated samples in a ball mill. The pre-treated mineral did exhibit a 32% higher grinding rate. The majority of the devices mentioned above were composed of a stationary vibrating surface opposing a passive rotating device that nips the particles in a gap that is smaller than the feed particle size. The ultrasonic vibration amplitude is about an order of magnitude smaller than the gap. The low processing capability attained with this design limits the method application, in spite of the very low energy consumption claimed by the authors. In this paper, new developments in applying ultrasound energy to grinding are presented. The design is based on a high-compression roller mill in which one of the rollers can be ultrasonically activated. In this way, the device retains its high performance, extending its ability to the efficient treatment of hard material. The two rollers apply mechanical stress to the material, facilitating the breakage process and coupling the active roller to the particles to be ground. To obtain a low specific energy consumption, the active roller was designed as a high-efficiency ultrasonic vibrator piezoelectrically driven at its third longitudinal resonant mode. To minimize any housing problems and to ensure a long operational life, the roller bearings were mounted at the nodal planes of the piezoelectric transducer. The breakage behaviour of the new device was investigated under several different loading conditions, and the actual effect of ultrasonic activation was evaluated. The optimal operational conditions were established by calculating the specific rate of breakage, a figure of merit that summarizes specific energy consumption, particle size distribution and material flow.
2. Description of the device In the design of a mill, the main problem to be overcome is the high energy and material consumption demanded by the size reduction of particles from a few millimetres to hundreds of micrometres. To broach the design of the new machine, and in view of the necessity for energy efficiency, the strategy adopted was the development of optimization criteria for the design of the ultrasonic components in the machine.
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
347
1. the power input line, which connects a motor to the gear train and is fitted with a power meter to measure the energy supplied to the rollers; 2. a torque sensor to monitor the roller mill during the test; 3. the ultrasonic roller mill; 4. a power generator with dynamic control that adjusts the frequency following the changes in the resonance of the ultrasonic system during operation [15]; 5. instrumentation to measure the applied ultrasonic power, the roller angular speed, and the instantaneous roller separation, allowing the stress over the shafts to be measured.
Fig. 1. Basic scheme of the ultrasonic roller and distribution of radial vibration.
The ultrasonic roller mill presented in this paper is a high-pressure roller mill in which at least one of the rollers is ultrasonically active. In order to obtain a high efficiency, the active roller was designed as a stepped rod driven at resonance in its third extensional mode (24.5 kHz) by means of a piezoelectric sandwich transducer ( Fig. 1). The nodal planes in the inner steps of the mechanical amplifier were used to mount the bearings. To assess the influence of the bearing housings, the impedance curve of the ultrasonic roller (the transducer coupled to the stepped rod ) was measured without constraints and inserted into the mill. As can be seen in Fig. 2, the behaviour of the ultrasonic roller in both conditions is similar except for the value of the impedance, which is greater for the roller inserted in the mill because of the inevitable losses. The radial vibrational amplitude distribution, measured for typical input powers with a laser vibrometer, is shown in Fig. 1. The distribution of the longitudinal vibration (j ) can easily be obtained from these values y by using the expression: j =sd(∂j /∂x), y x where s is the Poisson ratio, d is the rod diameter and ∂j /∂x is the radial displacement. x The active roller was mounted in front of the other passive roller forming a high-compression roller mill as shown schematically in Fig. 3. A linear bearing built into the bearing housing allows the rollers to be displaced with respect to each other. Displacement of the mobile roller is controlled through the use of a spring array that provides the required working stress. The two rollers can also be operated with a static gap. Gears transmit angular velocity to the roller system and are designed to obtain the same tangential speed on both rollers. In Fig. 3, the schematic drawing of the experimental set-up is shown. The system contains the following parts:
3. Results The ultrasonic roller mill was used to reduce the size of several substances under different operating conditions. To simulate the operating conditions in grinding plants, the system was always operated in a shock-feed state; the condition reached when the chamber formed by the two rollers is full of material, as opposed to the single particle feed state. After each test, the size distribution of the mill products was determined by screening. To ascertain the breakage pattern of the ultrasonic roller-mill, a series of grinding tests were conducted with different materials and in different operation conditions. The first test attempted to compare the grinding quality for the high-compression roller mill and the ultrasonic machine. The experiment was carried out by feeding the ultrasonic mill with different size distributions of quark, a brittle material, with and without ultrasonic activation. The results were very promising and showed that the grinding quality remained unchanged when the high-compression roller mill was ultrasonically activated. This conclusion is supported by Fig. 4, which shows the curves representing the size distribution of the product obtained with an ultrasonically activated machine at different input powers, and with the machine operating as a high-compression roller mill (0 W ), collapsing into one curve. The same results are also obtained if the product of the first grinding is fed to the mill for a second grinding process. For the second test of the ultrasonic grinding machine, copper mineral from Andina_Codelco was used. This is a very hard granite mineral that causes a high consumption of grinding media due to the high wear. Again, the curves shown in Fig. 5(a) for the ultrasonic grinding machine and Fig. 5(b), for the highcompression roller mill, show very similar size distributions. However, for the ultrasonically activated machine [Fig. 5(a)], the greater slope of the size distribution curves shows that a larger proportion of fine particles were produced. To assess the significance of these results on the
348
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
Fig. 2. Impedance diagrams of the ultrasonic roller measured without constrains (a) and inserted into the mill (b).
Fig. 3. Schematic representation of the experimental set-up to test the ultrasonic grinder.
energy consumption of the process, a further series of experiments were carried out on the mill, by applying different levels of ultrasonic power. The results, presented in Fig. 6, show that when ultrasound is applied, the total consumed power diminishes inversely to the ultrasonic power applied. Of course, this relation between the ultrasonic power applied and the total amount of energy used will be linear-inverse when the transducer is operated at its linear dynamic range. Fig. 6 shows that the total power needed by the process drops by approximately 15% for an applied ultrasonic power of approximately 100 W. This result suggests that it would be interesting to explore the application of even higher ultrasonic powers than that tested in this experiment. However there are two disadvantages for this approach. Firstly, for 100 W of applied
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
349
Fig. 4. Cumulative size distributions for quark ground in the ultrasonic roller mill with a fix gap setting of 150 mm. The angular speed of the rolls is 100 rpm.
power, the transducer begins to lose its linearity, and secondly, as the ultrasonic power applied increases, the material flow diminishes, making the process less attractive. These aspects will be treated in due course, when the work will focus on optimizing the whole process. Special tests to establish the mechanical behaviour and wear characteristics of the ultrasonic grinding machine were also carried out. Fig. 7(a) shows the reduction of torque with ultrasonic power applied, and in line with the results obtained previously, the behaviour of these variables is still linear. For the wear test, the ultrasonic machine was equ-
Fig. 5. (a) Size distribution of copper mineral ground in the ultrasonic actived high-compression roller mill. (b) Size distribution of copper mineral ground in the same device without ultrasonic activation. The angular speed of the rolls was 120 rpm.
ipped with two sets of rollers constructed with relatively soft steels. Wear tests were carried out for each one. For rollers built with SAE 1020 steel, the ultrasonic activation of the machine reduces material consumption by 54%, while for SAE 4340 steel, the material consumption is reduced by 38%. These results are shown in the Fig. 7(b). To obtain comparative results of the ultrasonic machine with the technology currently in use to grind hard materials, two equal amounts of copper minerals from Andina mining were ground to a powder to the same degree with the ultrasonic device and in a ball mill ( Fig. 8). It can be deduced from Fig. 8 that the size distributions in the ultrasonic machine and in the ball mills are
Fig. 6. Total power consumed vs. applied acoustic power for the same grinding task at 120 rpm.
350
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
the ultrasonic device are at rather low levels of applied acoustic powers, between 10 and 30 W, but even for an applied power of 400 W, the whole behaviour of the ultrasonic mill is still better.
4. Results analysis
Fig. 7. (a) Curve showing the torque diminution vs. acoustic applied power in the ultrasonic machine. (b) Curve for abrasive wear test. Mineral from ‘Minera Pudahuel’, 100% size — 6# Tyler.
approximately the same. However, the processing time was 40 min for the ball mill and only 1.85 min for the ultrasonic machine. In addition, the rate of energy consumption in the ultrasonic machine was 6.8 kWh/ton, and for the ball mill, this value increased to 20 kWh/ton. The specific breakage velocity is a very useful parameter with which to evaluate grinding technologies: it includes the flow of treated material, the size distribution and the specific use of energy [16 ]. Operationally, this may be considered as the figure of merit for different grinding technologies. In Fig. 9, curves for the specific breakage velocity at different levels of applied ultrasonic power are presented. The best operating conditions for
One of the most surprising features found during testing of the ultrasonic grinding machine was the important reduction in wear. This effect can be explained using a theoretical model to assess the effectiveness of ultrasound in decreasing the friction coefficient in rolling processes [17]. The friction coefficient changes in the ultrasonic machine because after the nip angle is reached, a compaction process is produced in the grinding material. The stressed material behaves like a solid, and the extensional vibrations of the active roller are normal to the material-sliding vector. The points of the active roller oscillate around their equilibrium position with a speed n, a result of which is a change in the direction of the tangential relative velocity, n of the rollers. In this 0 situation, the efficiency in the diminution in the wear, g (%), can be calculated with the expression [17] g (%)=1−
2
n s p En +n a s
S A 1−
P
p/2
0 dt
1
1+(n /n ) s a
B
, sin2(2pt/T )
where n is the sliding velocity, n is the vibration s a
Fig. 8. Size distribution for the same material ground with the ball mill and the ultrasonic roller mill until to obtain the same degree of finesse.
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
351
Fig. 9. Specific breakage velocity versus the acoustic applied power.
velocity, t is the integration time, and T is the wave period The last expression is a complete elliptical integral that can be solved by numerical methods. The solution for our experimental conditions is shown in Fig. 10. For an applied ultrasonic power of 100 W, the extensional vibration amplitude measured by optical means and averaged over the whole active roller is approximately 0.6 m/s. This value reasonably predicts the effectiveness of ultrasound in reducing of the wear of SAE 1020 steel. For the other steels, the fit is not so good, showing, unsurprisingly, that the wear also depends on the material properties. The experimental wear data point obtained would appear to agree closely with the predicted value. A more detailed experimental wear analysis, using a variety of
different steels and velocities, is beyond the scope of this paper. Finally, new tests were carried out to find the optimum vibrational displacement amplitude for the best ultrasonic grinding. In this way, the localized magnitudes that optimize the process can be established. These values will be critical to scale up the developed machine. The average displacement amplitude found for an ultrasonic applied power of 25 W was approximately 4 mm. It should be noted that the capacity of the laboratory ultrasonic grinding machine developed is 20 ton/day with an active roller of 200 mm in length.
5. Conclusions A new kind of ultrasonic grinding machine has been designed, constructed and tested. The results showed that the ultrasonic device produces high-quality ground products with a lower energy and material consumption than the conventional systems. Some of the localized magnitudes for scaling up the device were determined. Further research work to determine the fundamental parameters of the ultrasonic fracture of material has to be carried out in order to optimize the process. Ultrasonic grinding is a very promising new technology, as it may help improve the efficiency and to broaden the capabilities of conventional grinding systems.
Acknowledgement Fig. 10. Variation in effectiveness of ultrasound on contact friction versus the oscillatory speed amplitude in the rolls.
The authors wish to acknowledge the financial support from FONDEF Project D97T1011.
352
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
References [1] S. Bernotat, K. Scho¨nert, Ullmann’s Encyclopedia of Industrial Chemistry, fifth ed., VCH Verlagssesellschaft, Weinheim, 1988. Chapter 5. [2] M. Rhodes, Introduction to Particle Technology, Wiley, New York, 1988. [3] V.I. Revnitsev, in: E. Forssberg ( Ed.), XVI Int. Mineral Processing Congress, Elsevier Science, Amsterdam, 1988, pp. 93–116. [4] K. Scho¨nert, Int. J. Miner. Process. 22 (1988) 401–412. [5] H.K. Hutter, Eng. Mining J. 3 (2) (1970) 88–89. [6 ] B.K. Parekh, H.E. Epstein, W.M. Goldberger, Mining Eng. 36 (9) (1984) 1305–1309. ¨ ber die Mo¨glichkeit der zerkleinerung suspendierter [7] W. Ga¨rtner, U stoffe durch ultrashall, Acustica 3 (1953) 124–128. [8] W.J. Fleischhauer, C. Kro¨ger, Forschungsberichte des Landes Nordrhein-Westfallen No. 2081, (1969). [9] K.F. Graff, In: Ultrasonic Comminution, Ultrasonics International ’79 Proc., May, Graz, Austria (1979) 171–175.
[10] M.F. Leach, G.A. Rubin, Fragmentation of Rocks Under Ultrasonic Loading, Ultrasonic Symp. IEEE, (1988). [11] W.B. Tarpley Jr., P.L. Howard, G.R. Moulder, Efficient Ultrasonic Grinding: a New Technology for Micron-sized Coal, Quarterly Technical Progress Report No. 2, (1980). [12] T. Link, R.P. Killmeyer, Ultrasonic Comminution of Coal, PETC Coal Preparation Division Internal Report, (1988). [13] Y.C. Lo, P.J. Kientzler, in: S.K. Kawatra (Ed.), Comminution — Theory and Practice Symp., Society for Mining Metallurgy and Exploration, Littleton, 1992, pp. 645–659. Chapter 47. [14] J. Menacho, C. Yerkovic, L. Gaete-Garreto´n, Miner. Eng. 6 (1993) 6. [15] J. Gallego-Jua´rez, G. Rodrı´, J. San Emeterio, F. Montoya-Vitini, European Patent No. EP0450030, 1990. [16 ] L.G. Austin, R.R. Klimpel, P.T. Luckie, Process Engineering of Size Reduction: Ball Mill, AIME, New York, 1984. [17] V.P. Severenko, V.V. Klubovich, A.V. Stepanenco, Ultrasonic Rolling and Drawing of Metals, Consultants Bureau, New York, 1972.
Application of ultrasound in comminution L.F. Gaete-Garreto´n *, Y.P. Vargas-Herma´ndez, C. Velasquez-Lambert Laboratorio de Ultrasonidos, Depto. de Fı´sica USACH, Casilla 307, Santiago-2, Chile
Abstract This paper deals with a new technology for fine grinding of hard materials, based on a high-compression roller mill with ultrasonic capabilities. The machine was tested by producing fine powders from hard rocks, with and without ultrasonic activation, permitting the beneficial effects of ultrasound to be evaluated. The experimental set-up allows the following operational parameters to be measured: material flow, applied torque, angular velocity of the rollers, stress on the shafts, ultrasonic energy applied and the vibration amplitude and phase behaviour of the transducer roller. It is found that the application of ultrasonic energy diminishes the torque required, the stress over the shafts and the total energy consumed for the same grinding results. In addition, a reduction in the erosion of the grinding surfaces was found. The optimal value of the applied ultrasonic power was determined by measuring the specific rate of breakage, a parameter that refers to the energy consumed for the generation of 1 ton of material, for each size range. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Comminution; Grinding; High-compression ultrasonic mill; Size reduction
1. Introduction Comminution is defined as the mechanical breakdown of solids into smaller particles without changing the state of aggregation [1]. This process may be divided into two steps: crushing, which is the reduction of large pieces of material to a size suitable for grinding, and grinding itself, the reduction of crushed material to a powder. In order to produce a mechanical fracture, the material must be stressed beyond a certain critical limit. At present, three main stress mechanisms for particle size reductions are generally used: $ stress applied between two surfaces, applying normal and tangential forces to a single or many particles in the bulk material as in various crushers ( jaw crushers, overhead eccentric jaw crushers, impact crushers, giratory crushers, disk crushers, roll crushers, hammer crushers roller and ring-roller mills); $ stress applied at a single solid surface (surface– particle or particle–particle) such as in grinding media
* Corresponding author. Tel.: +56-2-7797-067; fax: +56-2-7797-067. E-mail address: [email protected] (L.F. Gaete-Garreto´n)
mills (tumbling mills, vibration mills, planetary mills, autogenous mills); $ stress applied by a carrier medium, such as in wet grinding. In conventional grinding machines, the energy that is actually needed for particle reduction is approximately 1% of the total energy applied to the machine. Despite the economic importance of comminution, its technology and efficiency have basically remained unchanged for many years, leading to a great demand for improvements [2,3]. Recent developments include the high-compression roller mill [4], a more efficient machine that introduces several advantages over the classical roller mill. It provides a better performance and is capable of achieving fine grindings from soft materials but is not useful for grinding hard materials due to excessive mechanical wear. The low efficiency of current grinding technologies is mainly due to the application of the stress to the whole material volume. This produces an excess of elastic energy in the load, an important part of which is finally dissipated as heat. Therefore, to increase the grinding process efficiency, it would be advisable to look for a way to apply the grinding energy only to the specific material to be fractured.
0041-624X/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S0 0 4 1 -6 2 4 X ( 9 9 ) 0 0 17 0 - 5
346
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
Some mechanisms to supply the breaking energy just in the grinding material have been explored previously. An example is the Snyder process, which uses a pressure discharge in a chamber filled with coarse material to produce fluidization and a grinding effect. Thermal stress, electrical discharge and electro-hydraulic effects have also been tried as alternative non-impact grinding methods, but have resulted in energy demands an order of magnitude higher than those needed by traditional technology to obtain the same product [5,6 ]. In the meeting of International Comminution Research Association, in Warsaw in 1993, the feasibility of the application of ultrasound energy to the grinding process was seen a viable avenue of exploration and research. Several different reasons were argued in the meeting to support such a proposition. One of the most convincing originated in the accepted fact that inside any material there are a number of inherent cracks and ultrasonic energy has the capacity to produce crack propagation from within the particle to its outer surface, producing an efficient fracture. In addition, in the case of mineral grinding, this method would be qualitatively better than the application of external stress, because the fractures would occur in the natural boundaries within the material compounds, precluding inter-crystalline breakage, a phenomenon that produces new substances, complicating the extraction process. In synthesis, the objective is the efficient transfer of ultrasonic energy from the source to the material to be ground. The fragmentation effect of ultrasonic fields in suspensions and in solids has been investigated since the fifties. Ga¨rtner [7] was probably the first researcher to have attempted using ultrasonic waves in the fragmentation of particles, obtaining poor results. Fleischhauer [8] used ultrasonic waves to fragment coal in suspension, while Graff [9] arranged many transducers in a multistage stamp mill. Quantitative information on the success of these attempts is very scarce. Leach [10] studied the fragmentation of resonant rocks samples fixed to the tip of an ultrasonic transducer, observing a preferred fracture at the nodes. Tarpley and Moulder [11] studied the effect of ultrasound in the grinding of coal under compression. These researchers designed two different machines. The first consisted of a nip roll mill that squeezes the coal grains between a roll and an ultrasonic active plate. Their reported values of energy consumption were as low as 3 kWh/ton to obtain a product 80% under 125 mm, much lower than 20 kWh/ton usually spent in hammer mills. The second machine was a conical device, activated through two opposing ultrasonic transducers with an Auger spiral to control the flux of material and to squeeze the coal against the ultrasonically active conical wall. According to the authors, this device offered a
production capacity of 136 kg h with an energy consumption of 5–7 kWh/ton. Later, in 1988 [12], using the same device, other researchers were unable to reproduce the previously reported results. In 1992, Lo and Kientzler [13] evaluated the Tarpley nip roller machine, using the device to grind mineral specimens. Their results showed energy consumptions similar to that of a ball mill, but not better. Menacho et al. [14] carried out grinding tests comparing standard mineral samples with ultrasonic pre-treated samples in a ball mill. The pre-treated mineral did exhibit a 32% higher grinding rate. The majority of the devices mentioned above were composed of a stationary vibrating surface opposing a passive rotating device that nips the particles in a gap that is smaller than the feed particle size. The ultrasonic vibration amplitude is about an order of magnitude smaller than the gap. The low processing capability attained with this design limits the method application, in spite of the very low energy consumption claimed by the authors. In this paper, new developments in applying ultrasound energy to grinding are presented. The design is based on a high-compression roller mill in which one of the rollers can be ultrasonically activated. In this way, the device retains its high performance, extending its ability to the efficient treatment of hard material. The two rollers apply mechanical stress to the material, facilitating the breakage process and coupling the active roller to the particles to be ground. To obtain a low specific energy consumption, the active roller was designed as a high-efficiency ultrasonic vibrator piezoelectrically driven at its third longitudinal resonant mode. To minimize any housing problems and to ensure a long operational life, the roller bearings were mounted at the nodal planes of the piezoelectric transducer. The breakage behaviour of the new device was investigated under several different loading conditions, and the actual effect of ultrasonic activation was evaluated. The optimal operational conditions were established by calculating the specific rate of breakage, a figure of merit that summarizes specific energy consumption, particle size distribution and material flow.
2. Description of the device In the design of a mill, the main problem to be overcome is the high energy and material consumption demanded by the size reduction of particles from a few millimetres to hundreds of micrometres. To broach the design of the new machine, and in view of the necessity for energy efficiency, the strategy adopted was the development of optimization criteria for the design of the ultrasonic components in the machine.
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
347
1. the power input line, which connects a motor to the gear train and is fitted with a power meter to measure the energy supplied to the rollers; 2. a torque sensor to monitor the roller mill during the test; 3. the ultrasonic roller mill; 4. a power generator with dynamic control that adjusts the frequency following the changes in the resonance of the ultrasonic system during operation [15]; 5. instrumentation to measure the applied ultrasonic power, the roller angular speed, and the instantaneous roller separation, allowing the stress over the shafts to be measured.
Fig. 1. Basic scheme of the ultrasonic roller and distribution of radial vibration.
The ultrasonic roller mill presented in this paper is a high-pressure roller mill in which at least one of the rollers is ultrasonically active. In order to obtain a high efficiency, the active roller was designed as a stepped rod driven at resonance in its third extensional mode (24.5 kHz) by means of a piezoelectric sandwich transducer ( Fig. 1). The nodal planes in the inner steps of the mechanical amplifier were used to mount the bearings. To assess the influence of the bearing housings, the impedance curve of the ultrasonic roller (the transducer coupled to the stepped rod ) was measured without constraints and inserted into the mill. As can be seen in Fig. 2, the behaviour of the ultrasonic roller in both conditions is similar except for the value of the impedance, which is greater for the roller inserted in the mill because of the inevitable losses. The radial vibrational amplitude distribution, measured for typical input powers with a laser vibrometer, is shown in Fig. 1. The distribution of the longitudinal vibration (j ) can easily be obtained from these values y by using the expression: j =sd(∂j /∂x), y x where s is the Poisson ratio, d is the rod diameter and ∂j /∂x is the radial displacement. x The active roller was mounted in front of the other passive roller forming a high-compression roller mill as shown schematically in Fig. 3. A linear bearing built into the bearing housing allows the rollers to be displaced with respect to each other. Displacement of the mobile roller is controlled through the use of a spring array that provides the required working stress. The two rollers can also be operated with a static gap. Gears transmit angular velocity to the roller system and are designed to obtain the same tangential speed on both rollers. In Fig. 3, the schematic drawing of the experimental set-up is shown. The system contains the following parts:
3. Results The ultrasonic roller mill was used to reduce the size of several substances under different operating conditions. To simulate the operating conditions in grinding plants, the system was always operated in a shock-feed state; the condition reached when the chamber formed by the two rollers is full of material, as opposed to the single particle feed state. After each test, the size distribution of the mill products was determined by screening. To ascertain the breakage pattern of the ultrasonic roller-mill, a series of grinding tests were conducted with different materials and in different operation conditions. The first test attempted to compare the grinding quality for the high-compression roller mill and the ultrasonic machine. The experiment was carried out by feeding the ultrasonic mill with different size distributions of quark, a brittle material, with and without ultrasonic activation. The results were very promising and showed that the grinding quality remained unchanged when the high-compression roller mill was ultrasonically activated. This conclusion is supported by Fig. 4, which shows the curves representing the size distribution of the product obtained with an ultrasonically activated machine at different input powers, and with the machine operating as a high-compression roller mill (0 W ), collapsing into one curve. The same results are also obtained if the product of the first grinding is fed to the mill for a second grinding process. For the second test of the ultrasonic grinding machine, copper mineral from Andina_Codelco was used. This is a very hard granite mineral that causes a high consumption of grinding media due to the high wear. Again, the curves shown in Fig. 5(a) for the ultrasonic grinding machine and Fig. 5(b), for the highcompression roller mill, show very similar size distributions. However, for the ultrasonically activated machine [Fig. 5(a)], the greater slope of the size distribution curves shows that a larger proportion of fine particles were produced. To assess the significance of these results on the
348
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
Fig. 2. Impedance diagrams of the ultrasonic roller measured without constrains (a) and inserted into the mill (b).
Fig. 3. Schematic representation of the experimental set-up to test the ultrasonic grinder.
energy consumption of the process, a further series of experiments were carried out on the mill, by applying different levels of ultrasonic power. The results, presented in Fig. 6, show that when ultrasound is applied, the total consumed power diminishes inversely to the ultrasonic power applied. Of course, this relation between the ultrasonic power applied and the total amount of energy used will be linear-inverse when the transducer is operated at its linear dynamic range. Fig. 6 shows that the total power needed by the process drops by approximately 15% for an applied ultrasonic power of approximately 100 W. This result suggests that it would be interesting to explore the application of even higher ultrasonic powers than that tested in this experiment. However there are two disadvantages for this approach. Firstly, for 100 W of applied
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
349
Fig. 4. Cumulative size distributions for quark ground in the ultrasonic roller mill with a fix gap setting of 150 mm. The angular speed of the rolls is 100 rpm.
power, the transducer begins to lose its linearity, and secondly, as the ultrasonic power applied increases, the material flow diminishes, making the process less attractive. These aspects will be treated in due course, when the work will focus on optimizing the whole process. Special tests to establish the mechanical behaviour and wear characteristics of the ultrasonic grinding machine were also carried out. Fig. 7(a) shows the reduction of torque with ultrasonic power applied, and in line with the results obtained previously, the behaviour of these variables is still linear. For the wear test, the ultrasonic machine was equ-
Fig. 5. (a) Size distribution of copper mineral ground in the ultrasonic actived high-compression roller mill. (b) Size distribution of copper mineral ground in the same device without ultrasonic activation. The angular speed of the rolls was 120 rpm.
ipped with two sets of rollers constructed with relatively soft steels. Wear tests were carried out for each one. For rollers built with SAE 1020 steel, the ultrasonic activation of the machine reduces material consumption by 54%, while for SAE 4340 steel, the material consumption is reduced by 38%. These results are shown in the Fig. 7(b). To obtain comparative results of the ultrasonic machine with the technology currently in use to grind hard materials, two equal amounts of copper minerals from Andina mining were ground to a powder to the same degree with the ultrasonic device and in a ball mill ( Fig. 8). It can be deduced from Fig. 8 that the size distributions in the ultrasonic machine and in the ball mills are
Fig. 6. Total power consumed vs. applied acoustic power for the same grinding task at 120 rpm.
350
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
the ultrasonic device are at rather low levels of applied acoustic powers, between 10 and 30 W, but even for an applied power of 400 W, the whole behaviour of the ultrasonic mill is still better.
4. Results analysis
Fig. 7. (a) Curve showing the torque diminution vs. acoustic applied power in the ultrasonic machine. (b) Curve for abrasive wear test. Mineral from ‘Minera Pudahuel’, 100% size — 6# Tyler.
approximately the same. However, the processing time was 40 min for the ball mill and only 1.85 min for the ultrasonic machine. In addition, the rate of energy consumption in the ultrasonic machine was 6.8 kWh/ton, and for the ball mill, this value increased to 20 kWh/ton. The specific breakage velocity is a very useful parameter with which to evaluate grinding technologies: it includes the flow of treated material, the size distribution and the specific use of energy [16 ]. Operationally, this may be considered as the figure of merit for different grinding technologies. In Fig. 9, curves for the specific breakage velocity at different levels of applied ultrasonic power are presented. The best operating conditions for
One of the most surprising features found during testing of the ultrasonic grinding machine was the important reduction in wear. This effect can be explained using a theoretical model to assess the effectiveness of ultrasound in decreasing the friction coefficient in rolling processes [17]. The friction coefficient changes in the ultrasonic machine because after the nip angle is reached, a compaction process is produced in the grinding material. The stressed material behaves like a solid, and the extensional vibrations of the active roller are normal to the material-sliding vector. The points of the active roller oscillate around their equilibrium position with a speed n, a result of which is a change in the direction of the tangential relative velocity, n of the rollers. In this 0 situation, the efficiency in the diminution in the wear, g (%), can be calculated with the expression [17] g (%)=1−
2
n s p En +n a s
S A 1−
P
p/2
0 dt
1
1+(n /n ) s a
B
, sin2(2pt/T )
where n is the sliding velocity, n is the vibration s a
Fig. 8. Size distribution for the same material ground with the ball mill and the ultrasonic roller mill until to obtain the same degree of finesse.
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
351
Fig. 9. Specific breakage velocity versus the acoustic applied power.
velocity, t is the integration time, and T is the wave period The last expression is a complete elliptical integral that can be solved by numerical methods. The solution for our experimental conditions is shown in Fig. 10. For an applied ultrasonic power of 100 W, the extensional vibration amplitude measured by optical means and averaged over the whole active roller is approximately 0.6 m/s. This value reasonably predicts the effectiveness of ultrasound in reducing of the wear of SAE 1020 steel. For the other steels, the fit is not so good, showing, unsurprisingly, that the wear also depends on the material properties. The experimental wear data point obtained would appear to agree closely with the predicted value. A more detailed experimental wear analysis, using a variety of
different steels and velocities, is beyond the scope of this paper. Finally, new tests were carried out to find the optimum vibrational displacement amplitude for the best ultrasonic grinding. In this way, the localized magnitudes that optimize the process can be established. These values will be critical to scale up the developed machine. The average displacement amplitude found for an ultrasonic applied power of 25 W was approximately 4 mm. It should be noted that the capacity of the laboratory ultrasonic grinding machine developed is 20 ton/day with an active roller of 200 mm in length.
5. Conclusions A new kind of ultrasonic grinding machine has been designed, constructed and tested. The results showed that the ultrasonic device produces high-quality ground products with a lower energy and material consumption than the conventional systems. Some of the localized magnitudes for scaling up the device were determined. Further research work to determine the fundamental parameters of the ultrasonic fracture of material has to be carried out in order to optimize the process. Ultrasonic grinding is a very promising new technology, as it may help improve the efficiency and to broaden the capabilities of conventional grinding systems.
Acknowledgement Fig. 10. Variation in effectiveness of ultrasound on contact friction versus the oscillatory speed amplitude in the rolls.
The authors wish to acknowledge the financial support from FONDEF Project D97T1011.
352
L.F. Gaete-Garreto´n et al. / Ultrasonics 38 (2000) 345–352
References [1] S. Bernotat, K. Scho¨nert, Ullmann’s Encyclopedia of Industrial Chemistry, fifth ed., VCH Verlagssesellschaft, Weinheim, 1988. Chapter 5. [2] M. Rhodes, Introduction to Particle Technology, Wiley, New York, 1988. [3] V.I. Revnitsev, in: E. Forssberg ( Ed.), XVI Int. Mineral Processing Congress, Elsevier Science, Amsterdam, 1988, pp. 93–116. [4] K. Scho¨nert, Int. J. Miner. Process. 22 (1988) 401–412. [5] H.K. Hutter, Eng. Mining J. 3 (2) (1970) 88–89. [6 ] B.K. Parekh, H.E. Epstein, W.M. Goldberger, Mining Eng. 36 (9) (1984) 1305–1309. ¨ ber die Mo¨glichkeit der zerkleinerung suspendierter [7] W. Ga¨rtner, U stoffe durch ultrashall, Acustica 3 (1953) 124–128. [8] W.J. Fleischhauer, C. Kro¨ger, Forschungsberichte des Landes Nordrhein-Westfallen No. 2081, (1969). [9] K.F. Graff, In: Ultrasonic Comminution, Ultrasonics International ’79 Proc., May, Graz, Austria (1979) 171–175.
[10] M.F. Leach, G.A. Rubin, Fragmentation of Rocks Under Ultrasonic Loading, Ultrasonic Symp. IEEE, (1988). [11] W.B. Tarpley Jr., P.L. Howard, G.R. Moulder, Efficient Ultrasonic Grinding: a New Technology for Micron-sized Coal, Quarterly Technical Progress Report No. 2, (1980). [12] T. Link, R.P. Killmeyer, Ultrasonic Comminution of Coal, PETC Coal Preparation Division Internal Report, (1988). [13] Y.C. Lo, P.J. Kientzler, in: S.K. Kawatra (Ed.), Comminution — Theory and Practice Symp., Society for Mining Metallurgy and Exploration, Littleton, 1992, pp. 645–659. Chapter 47. [14] J. Menacho, C. Yerkovic, L. Gaete-Garreto´n, Miner. Eng. 6 (1993) 6. [15] J. Gallego-Jua´rez, G. Rodrı´, J. San Emeterio, F. Montoya-Vitini, European Patent No. EP0450030, 1990. [16 ] L.G. Austin, R.R. Klimpel, P.T. Luckie, Process Engineering of Size Reduction: Ball Mill, AIME, New York, 1984. [17] V.P. Severenko, V.V. Klubovich, A.V. Stepanenco, Ultrasonic Rolling and Drawing of Metals, Consultants Bureau, New York, 1972.